SOML Read: Rethinking the Read Operation Granularity of 3D NAND SSDs

Chun Yi Liu, Jagadish B. Kotra, Myoungsoo Jung, Mahmut T. Kandemir, Chita R. Das

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

NAND-based solid-state disks (SSDs) are known for their superior random read/write performance due to the high degrees of multi-chip parallelism they exhibit. Currently, as the chip density increases dramatically, fewer 3D NAND chips are needed to build an SSD compared to the previous generation chips. As a result, SSDs can be made more compact. However, this decrease in the number of chips also results in reduced overall throughput, and prevents 3D NAND high density SSDs from being widely-adopted. We analyzed 600 storage workloads, and our analysis revealed that the small read operations suffer significant performance degradation due to reduced chip-level parallelism in newer 3D NAND SSDs. The main question is whether some of the inter-chip parallelism lost in these new SSDs (due to the reduced chip count) can be won back by enhancing intra-chip parallelism. Motivated by this question, we propose a novel SOML (Single-Operation-Multiple-Location) read operation, which can perform several small intra-chip read operations to different locations simultaneously, so that multiple requests can be serviced in parallel, thereby mitigating the parallelism-related bottlenecks. A corresponding SOML read scheduling algorithm is also proposed to fully utilize the SOML read. Our experimental results with various storage workloads indicate that, the SOML read-based SSD with 8 chips can outperform the baseline SSD with 16 chips.

Original languageEnglish
Title of host publicationASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems
PublisherAssociation for Computing Machinery
Pages955-969
Number of pages15
ISBN (Electronic)9781450362405
DOIs
Publication statusPublished - 2019 Apr 4
Event24th International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2019 - Providence, United States
Duration: 2019 Apr 132019 Apr 17

Publication series

NameInternational Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS

Conference

Conference24th International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2019
CountryUnited States
CityProvidence
Period19/4/1319/4/17

Fingerprint

Scheduling algorithms
Throughput
Degradation

All Science Journal Classification (ASJC) codes

  • Software
  • Information Systems
  • Hardware and Architecture

Cite this

Liu, C. Y., Kotra, J. B., Jung, M., Kandemir, M. T., & Das, C. R. (2019). SOML Read: Rethinking the Read Operation Granularity of 3D NAND SSDs. In ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems (pp. 955-969). (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS). Association for Computing Machinery. https://doi.org/10.1145/3297858.3304035
Liu, Chun Yi ; Kotra, Jagadish B. ; Jung, Myoungsoo ; Kandemir, Mahmut T. ; Das, Chita R. / SOML Read : Rethinking the Read Operation Granularity of 3D NAND SSDs. ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. Association for Computing Machinery, 2019. pp. 955-969 (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS).
@inproceedings{f5de827cdc234cdf9bf6645e64a168c8,
title = "SOML Read: Rethinking the Read Operation Granularity of 3D NAND SSDs",
abstract = "NAND-based solid-state disks (SSDs) are known for their superior random read/write performance due to the high degrees of multi-chip parallelism they exhibit. Currently, as the chip density increases dramatically, fewer 3D NAND chips are needed to build an SSD compared to the previous generation chips. As a result, SSDs can be made more compact. However, this decrease in the number of chips also results in reduced overall throughput, and prevents 3D NAND high density SSDs from being widely-adopted. We analyzed 600 storage workloads, and our analysis revealed that the small read operations suffer significant performance degradation due to reduced chip-level parallelism in newer 3D NAND SSDs. The main question is whether some of the inter-chip parallelism lost in these new SSDs (due to the reduced chip count) can be won back by enhancing intra-chip parallelism. Motivated by this question, we propose a novel SOML (Single-Operation-Multiple-Location) read operation, which can perform several small intra-chip read operations to different locations simultaneously, so that multiple requests can be serviced in parallel, thereby mitigating the parallelism-related bottlenecks. A corresponding SOML read scheduling algorithm is also proposed to fully utilize the SOML read. Our experimental results with various storage workloads indicate that, the SOML read-based SSD with 8 chips can outperform the baseline SSD with 16 chips.",
author = "Liu, {Chun Yi} and Kotra, {Jagadish B.} and Myoungsoo Jung and Kandemir, {Mahmut T.} and Das, {Chita R.}",
year = "2019",
month = "4",
day = "4",
doi = "10.1145/3297858.3304035",
language = "English",
series = "International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS",
publisher = "Association for Computing Machinery",
pages = "955--969",
booktitle = "ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems",

}

Liu, CY, Kotra, JB, Jung, M, Kandemir, MT & Das, CR 2019, SOML Read: Rethinking the Read Operation Granularity of 3D NAND SSDs. in ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS, Association for Computing Machinery, pp. 955-969, 24th International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2019, Providence, United States, 19/4/13. https://doi.org/10.1145/3297858.3304035

SOML Read : Rethinking the Read Operation Granularity of 3D NAND SSDs. / Liu, Chun Yi; Kotra, Jagadish B.; Jung, Myoungsoo; Kandemir, Mahmut T.; Das, Chita R.

ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. Association for Computing Machinery, 2019. p. 955-969 (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - SOML Read

T2 - Rethinking the Read Operation Granularity of 3D NAND SSDs

AU - Liu, Chun Yi

AU - Kotra, Jagadish B.

AU - Jung, Myoungsoo

AU - Kandemir, Mahmut T.

AU - Das, Chita R.

PY - 2019/4/4

Y1 - 2019/4/4

N2 - NAND-based solid-state disks (SSDs) are known for their superior random read/write performance due to the high degrees of multi-chip parallelism they exhibit. Currently, as the chip density increases dramatically, fewer 3D NAND chips are needed to build an SSD compared to the previous generation chips. As a result, SSDs can be made more compact. However, this decrease in the number of chips also results in reduced overall throughput, and prevents 3D NAND high density SSDs from being widely-adopted. We analyzed 600 storage workloads, and our analysis revealed that the small read operations suffer significant performance degradation due to reduced chip-level parallelism in newer 3D NAND SSDs. The main question is whether some of the inter-chip parallelism lost in these new SSDs (due to the reduced chip count) can be won back by enhancing intra-chip parallelism. Motivated by this question, we propose a novel SOML (Single-Operation-Multiple-Location) read operation, which can perform several small intra-chip read operations to different locations simultaneously, so that multiple requests can be serviced in parallel, thereby mitigating the parallelism-related bottlenecks. A corresponding SOML read scheduling algorithm is also proposed to fully utilize the SOML read. Our experimental results with various storage workloads indicate that, the SOML read-based SSD with 8 chips can outperform the baseline SSD with 16 chips.

AB - NAND-based solid-state disks (SSDs) are known for their superior random read/write performance due to the high degrees of multi-chip parallelism they exhibit. Currently, as the chip density increases dramatically, fewer 3D NAND chips are needed to build an SSD compared to the previous generation chips. As a result, SSDs can be made more compact. However, this decrease in the number of chips also results in reduced overall throughput, and prevents 3D NAND high density SSDs from being widely-adopted. We analyzed 600 storage workloads, and our analysis revealed that the small read operations suffer significant performance degradation due to reduced chip-level parallelism in newer 3D NAND SSDs. The main question is whether some of the inter-chip parallelism lost in these new SSDs (due to the reduced chip count) can be won back by enhancing intra-chip parallelism. Motivated by this question, we propose a novel SOML (Single-Operation-Multiple-Location) read operation, which can perform several small intra-chip read operations to different locations simultaneously, so that multiple requests can be serviced in parallel, thereby mitigating the parallelism-related bottlenecks. A corresponding SOML read scheduling algorithm is also proposed to fully utilize the SOML read. Our experimental results with various storage workloads indicate that, the SOML read-based SSD with 8 chips can outperform the baseline SSD with 16 chips.

UR - http://www.scopus.com/inward/record.url?scp=85064632284&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064632284&partnerID=8YFLogxK

U2 - 10.1145/3297858.3304035

DO - 10.1145/3297858.3304035

M3 - Conference contribution

AN - SCOPUS:85064632284

T3 - International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS

SP - 955

EP - 969

BT - ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems

PB - Association for Computing Machinery

ER -

Liu CY, Kotra JB, Jung M, Kandemir MT, Das CR. SOML Read: Rethinking the Read Operation Granularity of 3D NAND SSDs. In ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. Association for Computing Machinery. 2019. p. 955-969. (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS). https://doi.org/10.1145/3297858.3304035